Institutionen för Kultur och Kommunikation
Camilla Sahlström
Upper Secondary Students’ Assessment of Four Women Speaking Four Different Varieties of
English
Engelska C-uppsats
Datum/Termin: Höstterminen 05 Handledare: Thorsten Schröter Examinator: Solveig Granath
Abstract
Titel: Upper Secondary Students’ Assessment of Four Women Speaking Four Different Varieties of English
Författare: Sahlström, Camilla Engelska C, 2005
Abstract: Society exhibits a wide variety of different languages with various prominent features. At the same time as we honour diversity, however our civilisation is coloured with prejudice and preconceptions. Even if there is a rather liberal view on language use today, dialects and accents still carry positive and negative connotations for a majority of citizens. Research shows, that we are prejudiced and that we have predetermined ideas when it comes to certain language varieties.
In this study, I take up four varieties of Standard English: American, English, Australian and Scottish. I focus on the associations Swedish students make when it comes to these four language varieties and how this transforms into attitudes towards the speakers. A language attitude study is carried out by using a modified Matched Guise Test. I explain the difference between dialect and accent, as well as societal attitudes to language varieties and present some prominent linguists and their methods. Finally, I draw some conclusions by comparing my results to previous findings.
Nyckelord: Language attitude research, Standard English varieties, dialects and accents, speech communities, MGT.
1. Introduction ... 1
2. Background... 2
2.1 Varieties of Standard English ... 2
2.1.1 Modern Standard English... 2
2.1.2 Standard Scottish English... 3
2.1.3 Standard American English... 3
2.1.4 Standard Australian English... 5
2.2 Differences in Pronunciation... 5
2.2.1 Consonants ... 5
2.2.2 Vowels... 6
2.2.3 Stressed and unstressed syllables ... 7
2.3 Speech communities ... 7
2.3.1 Sociostructural determinants; standardization and vitality ... 8
2.4 Attitudes to dialects and accents in society ... 9
2.4.1 Pear... 10
2.4.2 The Matched Guise Technique ... 10
2.4.3 Lambert study... 11
2.4.4 Cheyne... 12
3. Material and Method ... 13
3.1 Designing a modified Matched Guise Test... 13
3.2 Questionnaire and background questions... 13
3.3 Classroom study ... 14
4. Analysis and Results... 15
4.1 The background questions... 15
5. Discussion and Conclusion ... 22
5.1 Discussion ... 22
5.2 Conclusion... 22
References ... 23
Appendix 1 ... 25
Appendix 2 ... 26
Appendix 3 ... 30
1. Introduction1
Virtually all of us engage in social interaction regularly and frequently. In this interaction, our language is our most important tool. It is not merely an instrument for exchanging information, however: it is a key to the relations with our fellow citizens. It is mainly through language that we establish who we are and who the interlocutor is. Language is thus more than writing and speaking; it carries underlying connotations that we ought to consider.
Having established the multiple interconnections of social situations and language choices, we now come on to consider how such choices are socially meaningful to others. In fact, even a single vowel or consonant sound, contrasting with others or with our expectations, can have evaluative repercussions for its utterer. (Giles &
Coupland 1991:32)
In a democratic society, people are supposedly equal. Still, they strive to find their place in the social hierarchy. The manner in which people speak partly reflects their position on the social ladder. Non-standard speech might indicate that the interlocutor is from a different geographical area or belonging to a certain speech community2 (Aitchison 1992:104). In order to make our information as comprehensible as possible, it is important to know our interlocutor’s status. Neither RP3 (Received Pronunciation) nor a broad dialect might be appropriate on all occasions; it may cause annoyance or confusion and sometimes prejudice appears. Surveys and investigations, by among others Giles and Coupland, show that we are prejudiced when it comes to certain dialects (Giles and Coupland 1991:32-33).
In this paper, I focus on the associations Swedish students make when they hear different varieties of Standard English4 and how this transforms into attitudes towards the speaker. In order to measure current student attitudes towards varieties of English, I will use a modified version of an indirect technique, the Matched Guise Technique (MGT).
1I would like to thank the students and their teacher, the four speakers and everyone else who has helped me to execute and conclude this survey; thank you all.
2 A group of people who consider that they speak the same language (e.g. regional, ethnic, class and occupational factors as well as gender can be determinants of a speech community).
3 RP- a British English accent favoured by the aristocracy and upper-middle-classes, a social accent (Trudgill 2000:8).
4The variety of English normally used in print and taught in school and to non-native speakers learning the language (Trudgill 2000:5-8).
2. Background
2.1 Varieties of Standard English
We all use different dialects, accents, intonation, stress and pace when talking. The term dialect refers to lexical as well as pronunciation differences between varieties of language, whereas accent refers solely to differences in pronunciation. It is often important to distinguish between the two, as for example in the context of the dialect known as Standard English. It is appropriate to consider it a dialect, since Standard English differs grammatically and lexically from other English varieties (Trudgill 2000:5-6).
2.1.1 Modern Standard English
Modern Standard English has evolved from various dialects in the south east of England, around the London area, and has developed since the fourteenth century (Ronowicz 2003:12;
Trudgill 2000:6). It was the language of the royals, the rich and the well educated that through commerce and the use of printing spread all over England.
Standard English is that variety of English which is usually used in print, and which is normally taught in schools and to non-native speakers learning the language. It is also the variety which is normally spoken by educated people and used in news broadcasts and other similar situations. (Trudgill 2000:5-6)
In England as in other countries, there are regional linguistic differences connected to various geographical areas (Yallop 1999:32). Since English is a global language, there are additional varieties like e.g. American, Australian and Scottish English (Trudgill 2000:5-8). We can perceive grammatical and lexical differences as well as characteristic pronunciations among them (Yallop 1999:26-43). One influential British accent of Standard English is RP, which, however is somewhat controversial since it evokes many ambiguous feelings connected with social class. RP was and still is the Queen’s English, spoken by many wealthy people and the aristocracy (Sharwood-Smith 1999:58-59). If one was able to master RP, one was seen as a member of an elite and this could be helpful when applying for a position. Today, the popularity of RP is on the decline, probably because it carries the negative connotation of being superior and posh (Trudgill 2000:194-195). Moreover, it has lost ground to Estuary English, which is a mix of Cockney5, a southeast working class accent, and RP (Yallop
5 Cockney, by the way, is an old word with roots in the 14th century. The original word was cokene, which meant cock egg. It was used to tell of a deformed egg – something highly likely if a cock laid it. Later, in the 16th
1999:32). “Many people blend in to their accent enough of RP to sound educated […] but not enough to be associated with the upper social class that RP traditionally represents”
(Sharwood-Smith 1999:60). English people speaking RP are often regarded as well-educated but somewhat snobbish (Sharwood-Smith 1999:59-60).
2.1.2 Standard Scottish English
Standard Scottish English is basically Standard English spoken with a Scottish accent which differs from most other varieties of English. Scottish English grammar and spelling follow the rules of the Oxford English Dictionary (OED). “However, there are some unique characteristics, mainly in the phonological and phonetic systems, many of which originate in the country’s two autochthonous languages, the Scottish Gaelic language and Lowland Scots”
(Wikipedia 2005). Standard Scottish English is spoken by educated, middle-class, urban Scots and the speech often conforms to the grammatical norms of the written standard, especially on formal occasions. Additionally, colourful Scottish idioms leave their mark on the language (Eagle 2001). This melodious variety of English carries positive connotations and Scottish people are generally seen as friendly, hospitable and warm people (Edwards 1982:23). The predominant linguistic varieties spoken in Scotland are Scottish English and Lowland Scots (Eagle 2001). There is, moreover, an additional variety in Scotland, Highland English, which differs greatly from Scottish English since Gaelic has influenced it (Wikipedia 2005).
2.1.3 Standard American English
Standard American English differs grammatically, lexically and phonetically from Standard English. Early settlers have contributed to the dialectal mixture and development of American English. For example, Swedish and Norwegian have been influential on the dialects around the Great Lakes, whereas in the Midland region, dialects from the British Isles were prominent (Ash 1999:202-203). As can be seen on map 1, the dialectal lines dividing the country are horizontal from left to right and, quite possibly, give an indication of how North America was settled by English speakers (Trudgill 2000:153).
century, the general meaning became “city folks” – they were considered soft and decadent by the rural people (Svartvik 1999:180).
Map 1. Modern North American dialect areas (Trudgill, P. 2000:154)
The pronunciation of English in the USA is noticeably different from British English (Yallop 1999:33) and shows in some respects more similarities with Scottish and regional British forms of speech than with RP. Regional diversity is not as marked in the USA as in Britain but the following dialectal regions can be distinguished: The West, the South, the Midland, North Central, North Inland, Eastern New England and New York City (Trudgill 2000:153- 154). The most commonly used accent in the U.S. is General American. Other characteristic varieties are Southern, North eastern and the sociolect African American Vernacular English.
Even if the Standard English spoken in England is considered as “The English language” and as such, very prestigious, American English is spoken by almost three times as many people.
The U.S. is the world leader in fields such as commerce, technology, and the media industry, thus being the number one representative of the English language. American English carries positive connotations due to prestige, power and media coverage. However, Americans are sometimes considered somewhat ostentatious and superficial as the American proverb “keep up with the Joneses”6 might imply (Ash 1999:205).
6 Striving to be richer or at least at the same social-level as the next-door neighbour.
2.1.4 Standard Australian English
There are three common varieties of Standard Australian English: Broad, General and Cultivated (Yallop 1999:33). Compared to the regional and social differences in America and in England, the varieties of pronunciation in Australia are not as obvious. Even though the English-speaking settlers established their characteristic pronunciation early, there seems to be “[…] an educational and cultural tradition of decrying local speech and admiring an (often highly idealized) English model” (Yallop 1999:32).
When foreigners think of Australian English, they usually think of the broad variety; as examples, I might mention Mick in the ‘Crocodile Dundee’7 movies and Steve ‘The Crocodile Hunter’ Irwin8. They speak something equivalent to the American Southern drawl in Louisiana and South Carolina, a kind of typical rural Australian with e.g. long diphthongs (Wikipedia 2005). Most Australians use General Australian, as the name indicates; it is characterised by shorter vowels than Broad Australian English. Finally, there is Cultivated Australian English, which is spoken only by a minority of the population and shares many similarities with the British RP. A majority of the public figures formerly used Cultivated Australian English, but today, it is considered posh and a bit snobbish (Wikipedia 2005).
Australian English carries for most people positive and friendly connotations (Ronowicz 1999:20).
2.2 Differences in Pronunciation
When it comes to the pronunciation of English in America, there are more similarities with Scottish and other regional British forms of speech than with RP (Yallop 1999:33). There is also less regional diversity in America and Australia than in Britain. The North eastern and South eastern parts of the United States feature the most easily perceivable exceptions and dialects in New England, New York and in the South have little in common with the variety called General American (Yallop 1999:34).
2.2.1 Consonants
“In general, consonants vary less than vowels, but there is a major division between those who pronounce the so-called ‘post-vocalic r’ and those who don’t” (Yallop 1999:34). In Australia and in RP, “[t]he /r/ standing after a vowel (post-vocalic) and either at the end of a
7 A comical movie about the stereotypical Australian.
8 An Australian zookeeper who uses the broad variety.
word or before a consonant is simply not pronounced” (Yallop 1999:34). For example, one does not hear the /r/ sound in car, card, four, fort, beer and stairs. This is also called the non- rhotic pronunciation. The New England and Southern varieties in America distinguish themselves by being non-rhotic, whereas General American is rhotic (Yallop 1999:34).
Scotland as whole and large parts of Western and South western England are considered rhotic as well (Yallop 1999:34). The non-rhotic accents perform a so-called /r/-linking, e.g. in connected speech, if the following word starts with a vowel, as in ‘a star is born’ and ‘four o’clock’. Sometimes this /r/-linking even occurs in situations where it does not seem to be called for, as in Camilla/r/ and Charles. This is quite common and at times criticized for being an example of bad pronunciation (Yallop 1999:34).
Another consonant that behaves differently in various situations is /t/. In General American, Australian and some regional varieties in England, /t/ is normally pronounced /d/ and usually deleted altogether after /n/ (Yallop 1999:34). In some places in England, like London, the word-medial or word-final /t/ is pronounced with, or as, a glottal stop. The words better and bet are pronounced /be’er/ and be’. In RP, the medial /t/ is neither voiced to /d/ nor replaced by a glottal stop as it can be in other varieties (Yallop 1999:35).
2.2.2 Vowels
“English has a relatively large number of different vowel sounds, including several diphthongs” (Yallop 1999:35-36). The diphthongs vary considerably across the English- speaking world and help to distinguish different varieties of pronunciation.
In RP, the diphthong illustrated by ‘so’ is quite distinctive, and is sometimes mimicked as if it began with a vowel like the short ‘e’ of ‘bet’ and then moved towards the vowel of ‘boot’. In Aus and NZ (especially Broad varieties) the diphthongs in ‘say’ and ‘so’ begin with the tongue relatively low in the mouth, so much so that others sometimes claim that ‘say’ sounds like ‘’sigh’ and ‘so’ sounds like ‘sow’. (Yallop 1999:35)
Actually, sow9 in Australian, as well as in some regional British and North American speech, usually begins with a low front vowel similar to that of sat and continues toward a vowel similar to that of boot (Yallop 1999:35). However, this is merely one example of the numerous aspects of vowel pronunciation that distinguish English speech varieties.
9 Female pig
Another interesting aspect of vowel pronunciation concerns the vowel /u/ of words like tune, duke, student and dune. It is typically a diphthong (sounding like ‘you’) in Australian and RP whereas in General American it is pronounced as a single vowel like toon, dook, stoodent and doone (Yallop 1999:35-36). The vowel /a/ of words like after, laugh, path, bath, pass and fast, which are typically long in RP and Australian, tend to shorten in General American into the same as the vowels of bat and bad (Yallop 1999:36).
2.2.3 Stressed and unstressed syllables
“The tendency of most varieties of spoken English is to make a major difference between stressed and unstressed syllables and to show vowel reduction in unstressed syllables” (Yallop 1999:36). When non-rhotic pronunciation and vowel reduction are combined, as in RP for instance, words with completely different endings like China, minor, and diner still end in the same indeterminate vowel or schwa (Yallop 1999:36). RP can actually exclude the penultimate vowel pronunciation completely if a word has prominent stress on the first syllable, as in the case with words like secretary and February. This results in words like
‘secretry’ and ‘Februry’ or even ‘Febyuri’. This type of vowel reduction is not applicable in General American (Yallop 1999:36).
In RP, the distinction between schwa and a short /i/ vowel in unstressed syllables remains.
“For example the words ‘carrot’, ‘chattered’, ‘fishers’ have schwa in the second (unstressed) syllable, while ‘rabbit’, ‘carpet’ ‘chatted’ commonly have a short vowel like that of ‘kit’ in the second syllable” (Yallop 1999:36). On the other hand, this type of distinction is not made in Australia; the words chatted and chattered are pronounced exactly alike, with a schwa in the second syllable (Yallop 1999:37). Note that neither RP nor Australian English pronounces the /r/ in words like chattered and fishers (Yallop 1999:37).
2.3 Speech communities
“The concept of speech communities is basic to an understanding of regional and social variation in language, otherwise known as dialects” (Clark, Eschholz & Rosa 1998:267). All people belong to at least one speech community, a community where its members speak the same language variety and obey the norms for its appropriate use. When changing speech community, e.g. when leaving home for work, the speech variety normally changes. The varieties spoken among co-workers often include terms that are inappropriate or even
unintelligible at home and vice versa. This phenomenon of modified speech is referred to as code switching and depends on age, gender, religious beliefs, or the grade of formality needed (Trudgill 2000:105-106, 201). In every society, the different degrees of influence exerted by certain social groups are reflected in language variation and in attitudes toward those variations. The language variety spoken is an indicator of social status and through code switching it is possible to change one’s current social grouping for a more advantageous one also known as social stratification (Trudgill 2000:24-25).
“Typically, the dominant group promotes its patterns of language use as the model required for social advancement; and use of a lower-prestige language, dialect or accent by minority group members reduces their opportunities for success in the society as a whole” (Ryan et al.
1982:1). Within the dominant group, which uses the standard variety, social classification is distinguishable through the speaker’s pronunciation, intonation or vocabulary.
In general, speech cues can be used by listeners to make inferences regarding an individual’s personal characteristics (e.g. age, sex, intelligence), social group membership (e.g. regional, ethnic, class, occupational), and psychological states (e.g. need for social approval, interest in continuing an interaction, anxiety, depression). (Ryan et al. 1982:2)
2.3.1 Sociostructural determinants; standardization and vitality
According to Ryan et al. (1982:3-5), there are two sociostructural determinants fundamental to the evolvement and expression of language attitudes, namely standardization and vitality.
“Whereas standardization represents the codification of the status quo, vitality more directly reflects the forces for shifts in language use and in symbolic values” (Ryan et al. 1982:3).
Fishman stated that “a language variety is said to be standardized if a set of norms defining
‘correct’ usage has been codified and accepted within a speech community” (quoted in Ryan et al.1982:3). In other words, for a language to be standardized and vital, it needs the approval and confirmation of society and its elite, the social institutions, such as the government, schools and the media. When a language becomes standardized, it functions as a norm for dictionaries, grammars and style manuals (Ryan et al. 1982:3).
Evidently, society decides whether a language becomes standardized or not. However, not every language has one standard variety or emphasizes standard variety use; for example, in Malaysia the inhabitants speak several different languages, or in Norway where “[s]ome
people do speak a form of Standard Norwegian, but the majority do not, whatever the social situation” (Trudgill 2000:197).
Vitality is, according to Ryan and Giles (1982:4-5), the second sociostructural determinant of language attitudes. The more a variety is used the more significant it is for the community members, the greater the vitality. “Thus, French, can be considered as more vital for Québécois [sic] French Canadians who use the language for all their daily interactions than for those French Americans who use it only in the home and with relatives” (Ryan et al.1982:4). The more those in control support a language variety, the higher the probability of current and future vitality. As Fishman says, “the status of a language variety rises and falls according to the range and importance of the symbolic functions it serves” (quoted in Ryan et al.1982:4).
However, it should be noted, “there are contexts in which increasing vitality occurs for non- standard varieties as well as situations in which standard varieties lose vitality” (Ryan et al.1982:5). As examples, the increased vitality of Catalan in Spain and the Vernacular Black English in the United States, has in fact additionally caused an increased standardization of these speech varieties (Ryan et al. 1982:5-6).
2.4 Attitudes to dialects and accents in society
Our multicultural society encompasses a wide variety of different languages. However, at the same time as we honour diversity, our civilisation is coloured by preconceptions and prejudice. Even if people have a rather liberal view on language use, and though no dialect is intrinsically better or worse than another (Edwards 1982:21-22), dialects and accents carry positive or negative connotations for the majority of citizens (i.e. trigger certain feelings and value judgements in people) (Shuy 1998:292-293). In the musical My Fair Lady, Professor Higgins embodies our prejudice when it comes to language varieties. Eliza is a poor woman who speaks the ‘ghastly’ cockney accent and whose life changes dramatically for the better as soon as she has learned to use the ‘proper’ English, or the variety known as RP. The musical thus illustrates how ability to acclimatise to community standards, by using the prestige type of a linguistic variety (which can be a standard or non-standard variety depending on the community), will surely enhance the chances of social acceptance and status (Ryan, Giles &
Sebastian 1982:1).
Language plays a crucial part in our social interaction and is central to everything we do (Clark et al. 1998:ix). When communicating, our interlocutor constantly puts the language we use under scrutiny, and we monitor the way someone else speaks (Trudgill 2000:1-2). We start to acquire our language as soon as we are born and “[l]anguage attitudes are viewed as an integral component of children’s developing communicative competence” (Ryan et al.
1982:15). We integrate prevalent attitudes from our home and society and make them our own. Four-year-old children can distinguish between different dialects and exhibit similar preferences toward language variation as the parents (Ryan et al. 1982:15).
2.4.1 Pear
In 1931, T. H. Pear conducted what is today considered a classic study on personality traits of certain voices heard on the radio (Giles & Coupland 1991:33). Pear’s goal was to find out whether there was any correlation between voice parameters and actual dispositional states.
The conclusion was that vocal features overlapped only slightly with actual personality traits.
Pear’s result was rather unspectacular, “as it revealed little in the way of consistent connections between speech and actual personality of the speaker” (Campbell-Kibler 2005:64), however, “[…] study after study has shown that there is a quite considerable social consensus among listener-judges about the stereotypical traits associated with voices other surveys were conducted that yielded interesting findings” (Giles & Coupland 1991:33).
2.4.2 The Matched Guise Technique
There are three techniques currently used in the study of language attitudes: content analysis, direct and indirect method (Ryan et al.1982:3). I will here concentrate on the third, and more particularly on the Matched Guise Technique (MGT). “The procedure is built on the assumption that speech style triggers certain social categorizations which will lead to a set of group-related trait inferences” (Giles & Coupland 1991:34). Listeners think that they are assessing, different speakers. What the listeners are unaware of is that it is the same speaker all the time, a person who is merely changing his/her dialect, thus the listener is actually assessing the speech variety.
[…] evaluations of language varieties – dialects and accents – do not reflect either linguistic or aesthetic quality per se, but rather are expressions of social convention and preference which, in turn, reflect an awareness of the status and prestige accorded to the speakers of these varieties. (Edwards 1982:21)
In order to expose value judgements made by listeners when hearing different speech and language varieties in Montreal, W. E. Lambert designed the Matched Guise Technique in 1960. After listening to (supposedly) different people speaking various dialects, the listener- judges are expected to assess and rate these dialects (Giles & Coupland 1991:34).
Additionally, the listener-judges are to provide the speakers with different personality traits, categorized by Lambert, “[…] for example, along dimensions to do with competence traits such as intelligence, ambition and confidence, and social attractiveness/integrity traits, such as sincerity, friendliness or generosity provided for them on a questionnaire” (Giles &
Coupland 1991:34). In general, the answers provided by the MGT reveal stereotypical social attitudes towards a certain language variety. “This area of study has tended to be referred to as
‘language-attitudes’ research” (Giles & Coupland 1991:34).
2.4.3 Lambert study
In the original Lambert, Giles and Picard study 1975, results showed that we are predisposed to judge speakers of our own speech group more favourably (Giles & Coupland 1991:34).
There is, however, one exception: if the other group speaks a standardized form of the language, we are more willing to give credit to their dialect or accent.
The most consistent finding is that the standard variety elicits attributions of competence and high social status from all social groups. While nonstandard regional and ethnic variants in Britain tend to be evaluated more favourably than the standard on characteristics related to integrity/social attractiveness and group solidarity […]. (Ryan et al. 1982:13)
Minority groups may actually downgrade their own varieties more than other groups would do (Giles & Coupland 1991:36-37). On one hand, the results from this study showed that minority groups might suffer from second-rate social prestige. On the other hand, in 1975, Lambert et al. carried out an additional study where Franco–American and Anglo-American students in Northern Maine were to rate (inter alia) the social class of recorded speakers.
“Various kinds of European, Canadian and local accented speech were represented, together with standard American English. Speakers were matched for objective social class. No differences at all emerged with regard to perceived status” (Giles & Powesland 1975:45). This lack of correlation between speech and social class indicates that even if “[t]he stigma attached to speakers of nonstandard dialects is well-documented”, it is not an absolute rule and exceptions do occur (Cross, DeVaney & Jones 2001:213).
In addition to these findings, Giles and Coupland (1991:40) point at how people change their conceptions of speech varieties throughout their lives:
[…] cross-sectional studies have shown children becoming gradually more socialized into accepting the evaluative norms of standardized speech, adolescents identifying more with local sociolinguistic ideals during their teens, and the elderly becoming seemingly more tolerant of non-standard variants in the speech of others. (Giles and Coupland 1991:40)
2.4.4 Cheyne
In a 1970 study, using the MGT, Cheyne examined attitudes towards Scottish and English regional accents. She found that in most cases, both Scottish and English raters had a higher opinion of the English speakers’ accents. There were however irregularitites: “some (Scottish) subjects rated male Scottish speakers more favourably than English ones on the personality dimensions suggesting ‘warmth’, and both groups of judges evaluated them as being more
‘friendly’” (Edwards 1982:23).
There have been some negative voices regarding the MGT’s validity, because of its alleged artificiality (Edwards 1982:22). However, the MGT seems to be one of the most applicable methods hitherto. When conducting this kind of survey, it is important to make sure that the testing conditions show conformity; otherwise, the results may be misleading. Both target groups and readers should be as homogeneous as possible, in order to ascertain the test’s validity.
3. Material and Method
3.1 Designing a modified Matched Guise Test
I was interested in studying upper secondary students’ attitudes towards different varieties of Standard English. To this end, I decided to employ a type of MGT test and asked the students to evaluate four native speakers after listening to audio recordings where the speakers read identical texts in their language variety. The speakers were from America, Australia, England and Scotland. I consciously tried to limit the speaker variables as much as possible; therefore, all four speakers are women with an academic background and residents of Sweden since a significant number of years. These women vary somewhat in ages: the speakers from America and Australia are between 40 and 50 years old, whereas the speakers from England and Scotland are between 60 and 65. As to the text the four native speakers were asked to read aloud, I chose the introduction to an article in The National Geographic (October issue 2005), concerning the worldwide threat of a bird-flu (H5N1) pandemic and its current repercussions in Asia, written by Tim Appenzeller (see appendix 1). I chose this text, firstly, because its content is of interest to us all, regardless of age, sex or occupation. Secondly, in order not to lose the respondents’ interest and concentration I did not want a text that was too long. After having surmounted occasional technical difficulties, I recorded the four native speakers as they were individually reading this text.
3.2 Questionnaire and background questions
I started out by formulating some background questions for the students (see appendix 2). I intended to gather information about their native languages, how they rated their knowledge of English, and when, where and to what extent they used English outside school.10
Subsequently, I made a questionnaire with the help of which the students could assess the four different speakers by adjectives and nouns describing personality characteristics. I listed twenty-two character traits, ranging from ambition and arrogance to trustworthiness and
10 Since we sometimes have difficulties understanding how we acquire knowledge and, furthermore, problems expressing our notions in this respect, I additionally consulted a survey from the National Agency for Education, on how young people learn English (Skolverket 2005). Apparently, in addition to school, children acquire English from TV, movies, music, the Internet and computer games, reading and interaction with the English- speaking world.
wealth, and then asked the students to assess the speakers according to these. I intended to find out whether the respondents’ assessments were influenced by the different language varieties. Additionally, I asked them to guess the speakers’ age and nationality. Since oddities and mistakes occur frequently in the design of tests I conducted a pilot study. I chose three persons for this pilot and quite soon, I understood that I had to explain some of my questions more clearly.
3.3 Classroom study
In the crucial phase of this study, three upper secondary classes listened to the four speakers and answered the questions. Without the absentees and latecomers, the number of students in my survey added up to forty-eight, seventeen of who were male and thirty-one female, and most were native Swedish speakers. My initial thought was to see if there was any significant correlation between gender and the assessment of the speakers, but I strongly believe that in order to get a valid result in this respect, more respondents are needed. I actually did look into the matter (see appendix 3), but this did not seem to yield meaningful and interesting findings.
The first day, two of the classes participated. I started by going through the questionnaire with the students and then asked them if there were any questions. After giving some of them additional information, the students filled out the two pages of background questions. When this first task was concluded, I informed them about the next four questionnaires (i.e. one for each speaker) and let them know that the speakers were about to read identical texts. I then played speakers 1-4 in succession, with an interval between each so that the respondents would have the time to answer the questions connected to each speaker and I would be able to answer additional questions concerning the study. The duration of the entire study was about thirty minutes per class. The second day, the final part of the study was conducted in the same manner as the first two. These students were somewhat quicker, though, and had finished the survey in twenty minutes.
The last two questions of each speaker questionnaire concerned the speakers’ age and nationalities and were the only ones that could be answered with a right or wrong answer. The respondents were interested in finding out if they had guessed the correct answers. When the speakers’ actual ages were revealed, the respondents were quite surprised, since many of them had estimated the speakers as significantly younger. In addition, when it comes to the
speakers’ nationality, the respondents were often quite astonished when they heard that they had guessed wrong, because they seemed quite confident when giving their answers.
4. Analysis and Results
4.1 The background questions
In this survey, there were seventeen male and thirty-one female respondents, all from the three different levels of upper-secondary school and aged between 16 and 19 years. The distribution can be seen in figure 1 below. The majority of students were female. Why this is the case I cannot tell; however, more girls than boys continue with their studies after compulsory school (Engqvist 2005). Additionally, girls seem more inclined to continue with language studies (Engqvist 2005).
0 2 4 6 8 10 12 14
16 17 18 19
Age
Number
Male Female
Figure 1. Respondents’ age and sex.
The 48 respondents primarily had Swedish as their native language (see figure 2).
0 5 10 15 20 25 30
Swedish Finish Kurdish Estonian French Language
Number Male
Female
Figure 2. Respondents’ native language.
A majority of the respondents, 75% do not use English at home. This fact is not surprising since 87.5% are Swedish native speakers.
0 5 10 15 20 25
Yes No
Number
Male Female
Figure 3. Respondents who use English at home.
Twelve respondents, 2 male and 10 female, indicated that they used English at home.
However, only three of them answered that they use English more than 1-5% (see figure 4).
The two respondents who answered that they use English 26-30% and more at home are non- native Swedes.
0 1 2 3 4 5 6 7 8
1-5 6-10 11- 15
16- 20
21- 25
26- 30
31- 35
36- 40
41- 45
46- 50
51+
%
Number
Male Female
Figure 4. The extent of English respondents claim to use at home.
Almost 50% of the respondents state that they do not use English on any other occasion than in school and at home, 25% say that they use English while being abroad, approximately 20%
claim they use it with friends and relatives and 6% say they use English on the internet (see figure 5).
0 2 4 6 8 10 12 14 16
With friends /relatives
On
the Internet
Abroad
On n o o
the r oc
casions
Number
Male Female
Figure 5. Additional occasions when respondents may use English.
In this study, the female respondents seem to have a lower opinion of their knowledge of English than the male respondents do, as can bee seen in figure 6. Since I have no knowledge of the respondents’ grades, I am in no position to draw conclusions from these perceptions.
0 2 4 6 8 10 12 14 16
1 2 3 4 5
rated value
number
Male Female
Figure 6. Respondents’ rating of their own knowledge of English (1=not good, 5=very good).
The reason that I asked questions about the respondents’ nationality and the amount of time spent speaking or otherwise using English was that I intended to find out whether some respondents could more easily establish the speakers’ nationalities and/or be somewhat biased toward a certain nationality. As it turned out, this was not the case. The non-native Swedish respondents’ success in identifying speaker nationality was similar to that of the native Swedes.
4.2 Assessments by the students
Generally speaking, the respondents tended to judge the speakers younger than their actual age (see table, figure I).
The American speaker (1) was in the 41-50 age range. Only 5 of the students guessed this; the majority (42 out of 48) perceived her younger.
The English speaker (2) was in the 51-60 age range. Only 6 of the students guessed this; the majority (38 out of 48) perceived her younger.
The Australian speaker (3) was in the 41-50 age range. None of the students guessed this correctly; they all perceived her younger.
The Scottish speaker (4) was in the 61 and above age range. Only 19 of the students guessed this; the majority (29 out of 48) perceived her younger.
Even if the respondents’ guesses were generally too low, they managed, more or less, to order the speakers correctly. It should be pointed out, that the guesses may have influenced the results, e.g. a respondent who considers a certain speaker younger than the other speakers may identify him/herself with this ‘younger’ speaker and give her attributes according to the perceived data.
Table I. Estimated speaker age (the figures in bold face are the amount of respondents stimating correct speaker age).
figure 7). If this is anything to go by, and I think it is, the tudents have a rather good knowledge of the different varieties of Standard English. I should
Estimated age
Speakers age in years 20-30 31-40 41-50 51-60 61-
American 20 22 5 1 0
English 2 17 19 6 4
Australian 44 4 0 0 0
Scottish 4 5 10 10 19
e
Even though the respondents rated their knowledge of English rather low, many students got the speaker nationality right (see
s
add that the other category covers “I do not know” answers and suggestions meant to be humorous, such as “Norway”, “Åmål” etc.
0 5 10 15 20 25 30 35
ber Speaker 1 American
Speaker 2 English
Am erica
England Australia
Scotland Canad
a Irelan
d
New Zealand
Wale s
Other
Nationality
Num
Speaker 3 Australian Speaker 4 Scottish
Figure 7. Perceived nationality.
0 0,5 1 1,5 2 2,5 3 3,5 4
Tra its
Am bitious
ArrogantArticulat e
Benevolent Cond
escendingConfident Econom
ical Educated
Exper ienced
Friendly Funny Generous
Intelligent Lonely Loy
al
Organized Relax
ed Shy Sociable
Strict Tru
stworthy Wea
lthy
Speaker 1 American Speaker 2 English Speaker 3 Australian Speaker 4 Scottish
Figure 8. Assessment of speakers (where rated value 0=”do not agree at all” and 5=”agree completely”).
The average rating per each trait and speaker can be seen in figure 8. I will here point out the most prominent traits for each of the speakers, both the highest rated value as well as the lowest.
Speaker 1, the American speaker scored the highest, compared to the other speakers, on the traits ambitious, articulate, benevolent, educated, friendly, organized and trustworthy. She scored the lowest on the traits arrogant, condescending, funny, generous, sociable and wealthy. According to this assessment, speaker number 1 seems to be considered a well- structured woman who is ambitious, articulate, organized and educated. Furthermore, she seems to be a warm person, since the respondents viewed her as the most friendly and trustworthy speaker. Additionally, the American speaker seems the least arrogant and condescending of the four speakers. This correlates to the stereotypical view of Americans as professional, but she was not assessed as being pretentious.
Speaker 2, the English speaker, was seen as the most economical person, which corresponds nicely with the fact that she is perceived as wealthy. Snobbish and well-educated are
stereotypical features of an English person (see 2.1.1), so this seems to correlate with the respondents’ views of speaker 2 as she was considered as intelligent, strict, not friendly, condescending, economical and wealthy.
Speaker 3, the Australian speaker, scored the highest, compared to the other speakers, on the traits arrogant and shy11. However, according to the respondents she does not seem to be trustworthy, economical, intelligent or experienced. Speaker 3 received the lowest rating among the four speakers on the traits articulate, benevolent, confident, economical, educated (same value as the Scottish speaker though), experienced, intelligent, lonely, loyal, organized, relaxed, strict and trustworthy. This does not correspond to the previous findings of stereotypical Australian traits (see 2.1.4), which indicate that Australians are perceived as positive and friendly. I should mention that speaker 3 was perceived as much younger compared to the other speakers; moreover, she spoke at a considerably faster pace than the others did. This might explain why she was rated rather unfavourably.
Finally, speaker 4, the Scottish speaker, was rated as most experienced. The respondents viewed her as confident, experienced and relaxed. She was furthermore believed to be funny, loyal and somewhat less educated (together with the Australian speaker) than the other speakers. This corresponds very well with the stereotypical Scottish traits warm, friendly, hospitable but maybe not so educated (see 2.1.2). I should add that the Scottish speaker used a perceivably slower pace than the other speakers when reading the text, and this might have influenced the respondents’ ratings.
In section 2.1, I gave some examples of stereotypical ideas of various nationalities (in the English speaking world): Americans are thought of as boastful, English people are considered polite but snobbish, Australians are looked upon as relaxed and friendly, and finally, Scottish people are regarded as warm and friendly. Even if the findings from my study of Swedish upper secondary students’ attitudes on language varieties fail to some extent (the case of the Australian speaker) to correlate with these earlier mentioned stereotypical ideas, the majority of the traits assessed correspond with conventional ideas of people speaking a certain language variety. Additionally, this material indicates that we have predetermined notions of certain language varieties and it has been the purpose of this study to test this.
11 Shy people can be perceived as arrogant.
5. Discussion and Conclusion
5.1 Discussion
When writing this paper, I have realized how important it is to be consistent. In the initial voice uptake, I failed to realize that the pace of the four native speakers’ reading differed a lot.
This is something to take into consideration since it may have affected the respondents’ views of, in particular, the Australian speaker. I would, furthermore like to point out that the text of little Ngoan is a very serious piece of journalism and can have been the reason for the rather low rating on the speakers’ funniness.
5.2 Conclusion
In this paper, I have tried to elucidate some of the questions concerning dialectal prejudice, and in particular, whether Swedish upper secondary students are prejudiced against certain dialects or not. Moreover, I have presented some of the more prominent names in the
‘language attitude research’ field and parts of their work. In order to find out if Swedish upper secondary respondents show prejudice when they hear different language varieties, I decided to carry out an attitude test on my own.
If the assessments by the respondents are valid and are a representation of a certain part of society, then the answer is that we are prejudiced. We share certain ideas with our families and members of our speech community, and to a certain extent on a national, European or even global level. Our language variety is important and equally important is the manner with which we assess language varieties; it reveals which speech community we belong to, and is a part of who we are. Is there a way to identify underlying sources and factors that influence our assessments? Yes, I think there is. As Ryan et al. (1982:15-16) point out; children acquire their prejudice as they acquire language, with their breast milk, figuratively speaking. If we assume that we acquire our attitudes towards foreign language varieties in the same way as our first language (e.g. via family, friends and media exposure) then there are ways to identify these judgemental factors and preset ideas. We are a product of ‘our’ society and prejudice is a part of our lives.
References
Ash, S. 1999.The United States of America-The Land of Opportunity. In: Ronowicz, E. &
Yallop, C. (eds.). 197-264.
Campbell-Kibler, K. 2005. Listener Perceptions of Sociolinguistic Variables: The Case of (Ing). Stanford. CA: Stanford University Press.
Clark, V. P., Eschholz, P. A. and Rosa, A. F. 1998. (eds.). Language: Readings in Language and Culture. 6th ed. Boston. MA: Bedford/St. Martin’s.
Cross, J. B., De Vaney, T. and Jones, G. 2001. Pre-Service Teacher Attitudes toward Differing Dialects. Linguistics and Education 12 (4):211-227.
Edwards, J. R. 1982. Language Attitudes and Their Implications among English Speakers. In:
Ryan, E. B. and Giles, H. (eds.). 20-33.
Eagle, A. 2001. Wir Ain Leid. [Online] Available at
http://www.scots-online.org/grammar/sse.htm (accessed November 7, 2005).
Giles, H and Coupland, N. 1991. Language: Contexts and Consequences. Milton Keynes:
Open University Press.
Giles, H. and Powesland, P. F. 1975. Speech Style and Social Evaluation. London: Academic Press.
Ronowicz, E. 1999. Introduction. In: Ronowicz, E. and Yallop, C. (eds.). 1-25.
Ronowicz, E. and Yallop, C. (eds.). 1999. English: One Language, Different Cultures.
London: Cassell.
Ryan, E. B. and Giles, H. (eds.). 1982. Attitudes towards Language Variation: Social and Applied Contexts. London: Edward Arnold.
Ryan, E. B., Giles, H. and Sebastian, J. 1982. An Integrative Perspective for the Study of Attitudes Towards Language Variation. In: Ryan, E. B. and Giles, H. (eds.).1-19.
Sharwood-Smith, M. 1999. British Shibboleths. In: Ronowicz, E. and Yallop, C. (eds.). 46- 82.
Shuy, R. W. 1998. Dialects how they differ. In: Clark, V. P., Eschholz, P. A. and Rosa, A. F.
(eds.). 292-312.
Skolverket. 2005. Slutbetygen I grundskolan kvar på samma nivå. [Online] Available at http://www.skolverket.se/sb/d/204/a/5253;jsessionid=81E7422B3D22EE9957D3F04C60D7A D01(accessed December 15, 2005).
Skolverket. 2005. Upper secondary, English. [Online] Available at
http://www3.skolverket.se/ki03/front.aspx?sprak=EN&ar=0506&infotyp=8&skolform=21&id
=EN&extraId=(accessed December 5, 2005).
Svartvik, J. 1999. Engelska – öspråk, världsspråk, trendspråk. Falun: Nordstedts.
Trudgill, P. 2000. Sociolinguistics: An Introduction to Language and Society. 4th ed.
Harmondsworth: Penguin.
Wikipedia. 2005. Scottish English. [Online] Available at
http://en.wikipedia.org/wiki/Scottish English (accessed December 4, 2005).
Wikipedia. 2005. Australian English. [Online] Available at
http://en.wikipedia.org/wiki/Australian English (accessed November 7, 2005).
Yallop, C. 1999. English around the World. In: Ronowicz, E. and Yallop, C. (eds.).26-45.
Appendix 112
“Little Ngoan”, the text read by the four speakers
“Little Ngoan was buried behind her parent’s hut three weeks ago. Her grave, a bulky concrete tomb like others dotting the Vietnamese countryside, rests on high ground between a fishpond and yellow-green rice fields. At one end her family laid out her cherished possessions: a doll’s chair, a collection of shells, plastic sandals. They painted her tomb powder blue.
While Ngoan’s parents are off helping with the rice harvest, other relatives share their memories. ‘She was so small, just ten years old,’ says her grandmother, sitting on a hammock.
‘She was very gentle and a good student. If you look at her older sister’-the 17-year-old hangs back shyly-‘you can imagine what she was like.’ Ngoan’s grandfather, silent with grief, lights a stick of incense at her grave. The loss of a beloved child has hit this family hard. But ordinarily, the wider world would pay little attention to a child’s death from infectious disease in this remote corner of Vietnam’s Mekong Delta. Old scourges like dengue fever and typhoid still take a toll here, and HIV/AIDS is on the rise.
Yet Ngoan’s death and more than 50 others in Southeast Asia over the past two years have raised alarms worldwide. Affected countries are struggling to take action; other nations are sending aid and advisers while stockpiling drugs and developing vaccines at home. And scientists have stepped up their research into the fateful traffic of disease between animal and people. Why? Because Ngoan died of the flu.”
Tim Appenzeller is the author of the above text. It has been taken from the October 2005 issue of the National Geographic, and the four readings of it can be found on the enclosed cd or alternatively on an audio file.
However, there is one alteration of the text; the name Ngoan has been changed to Joan, in order to facilitate for the readers.
12 A CD or alternatively an audio file, with the four native English speakers reading the text will be enclosed with this paper.
Appendix 2
Background questions for students
Thank you all very much for participating in this survey.
Please answer these questions as truthfully as possible. Go through the entire form before answering and if there are any questions that you want clarified, please raise your hand and I will do my best to explain.
1. Are you male___ or female___?
2. How old are you? ___
3. Which language(s) would you consider your native language(s)? _____________________
___________________________________________________________________________
4. Do you speak English at home? ___
5. If yes, how much? (a possible answer could be, for example, “ca 10% of the time”) ___________________________________________________________________________
6. Do you speak English on other occasions than in school? ___
7. If yes, where?
_______________________________________________________________
8. How often would you say that you speak English per week, except from in school?
(a possible answer could be, for example, “never, twice a day, maybe five times a week or only a couple of weeks per year, when on holiday”)
___________________________________________________________________________
9. On a scale from 1 to 5 (where 1 is a very low level and 5 a very high one), how would you rate your knowledge of English? ____
Assessment of speakers
Speaker number 1
Please rate the following questions 1-5 according to a scale where 1= I do not agree and 5= I fully agree.
Do you think the speaker seems…?
1. arrogant (självsäker och nedlåtande) ___
2. sociable (sällskaplig) ___
3. friendly (vänlig) ___
4. loyal (solidarisk, trofast) ___
5. generous (generös, storsint) ___
6. trustworthy (trovärdig, pålitlig) ___
7. benevolent (välvillig) ___
8. funny (rolig) ___
9. relaxed (avspänd, avslappnad, lugn) ___
10. lonely (ensam, övergiven) ___
11. shy (blyg) ___
12. strict (sträng, hård, noggrann, exakt) ___
13. intelligent (intelligent, begåvad) ___
14. wealthy (förmögen, rik) ___
15. ambitious (ambitiös, arbetsam) ___
16. educated (bildad, kultiverad) ___
17. confident (säker, trygg, självsäker) ___
18. economical (ekonomisk, sparsam) ___
19. articulate (vältalig) ___
20. organized (organiserad, systematisk) ___
21. experienced (erfaren, rutinerad, van) ___
22. condescending (nedlåtande) ___
Finally,
How old do you think the speaker is? ______________
What is the speaker’s nationality (an English speaking country)? _________________
Questions for the four speakers
RECORDING ENGLISH VARIETIES
Speaker
Nationality:
From what region do you come/ what region has coloured your language the most?
In your opinion, what dialect do you speak?
Will you use your original dialect when recording?
If not, what dialect will you use?
Age: - 30__ 30-40__ 40-50__ 50-60__ 60 -__
How long have you been a Swedish resident?
Appendix 3
Speaker 1 American. Number of students rating each trait
Female Male Sum Female & Male 1=I do not agree - 5=I fully
agree
1=I do not agree - 5=I fully agree
1=I do not agree - 5=I fully agree
Traits 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Ambitious 4 4 9 10 4 2 4 8 1 2 6 8 17 11 6 Arrogant 15 8 3 4 1 7 8 1 0 1 22 16 4 4 2 Articulate 4 3 9 12 3 1 0 5 7 4 5 3 14 19 7 Benevolent 7 5 11 6 2 1 5 1 8 2 8 10 12 14 4 Condescending 10 13 4 2 2 6 5 3 2 1 16 18 7 4 3 Confident 2 15 9 4 1 1 5 6 3 0 3 20 15 7 1 Economical 3 3 14 8 3 1 3 5 8 0 4 6 19 16 3 Educated 2 4 9 12 4 1 3 5 6 2 3 7 14 18 6 Experienced 4 6 15 2 4 1 2 7 5 2 5 8 22 7 6 Friendly 2 11 14 4 0 2 4 5 5 1 4 15 19 9 1 Funny 29 2 0 0 0 16 0 0 0 1 45 2 0 0 1 Generous 10 11 8 2 0 5 4 5 2 1 15 15 13 4 1 Intelligent 3 5 13 9 1 2 3 7 4 1 5 8 20 13 2
Lonely 4 5 13 7 2 2 2 3 5 5 6 7 16 12 7
Loyal 2 10 11 8 0 1 2 9 5 0 3 12 20 13 0 Organized 3 5 8 10 5 1 1 3 8 4 4 6 11 18 9 Relaxed 6 7 10 5 3 4 1 4 5 3 10 8 14 10 6 Shy 9 6 7 6 3 2 6 7 0 2 11 12 14 6 5 Sociable 10 15 6 0 0 8 5 2 2 0 18 20 8 2 0
Strict 9 10 5 4 3 9 2 3 1 2 18 12 8 5 5
Trustworthy 2 5 8 12 4 1 2 10 3 1 3 7 18 15 5 Wealthy 9 15 6 1 0 5 7 4 1 0 14 22 10 2 0
